Thursday, January 5, 2012

JKH on MMT and Risk


This post brings together several disparately placed comments on the same topic for convenient reference, and for those that may not have seen the comments.

Initially, JKH commented on a post by Perry Mehrling at The Money View  concerning the recent article at The Economist on heterodoxy that mentions MMT. I posted both here. Hugh Heden asked for clarification here. JKH responded in the following comment. (I've adjusted the paragraphing with more white space for easier reading.)

JKH said... 

I made that comment very quickly so it was a bit sloppy.

First, I’m thinking of QE in the broadest sense – being an expansion of reserves on the right hand side of the balance sheet and an expansion of assets on the left.

Thinking about risk in the broadest terms, you have such categories as credit risk, liquidity risk, interest rate risk, and foreign exchange risk. All of these elements are involved one way or another when the Fed intervenes so broadly in the market. Some of the earlier programs involved more credit risk for example.

In the case of their treasury purchases, there is certainly an element of interest rate risk. People may not think there’s much interest rate risk these days due to the environment, but it exists as a risk, and different portfolio managers have different views on it. Bill Gross was famously wrong on his call of course.

One of the interesting things about interest rate risk in this environment is that marked to market risk (or present value risk) increases dramatically, the lower rates go, for a given absolute move in interest rates. That’s because lower rates increase the “duration” measure of bond risk. The low rate of discount used in the PV formula makes the longer term cash flows of a bonds super sensitive to very small interest rate moves.

Anyway, you can assess the amount of interest rate risk that’s been taken out of the market by doing a detailed analysis of what the interest rate risk might be on the Fed’s balance sheet, not only in present value terms, but in running interest margin terms. It’s a very complicated exercise and I won’t get into it here.

But basically you have the Fed now running a massive “short funding” position on its balance sheet, due to the longer duration of its assets against the immediate interest rate sensitivity of many of its liabilities – i. e. the reserves that are priced essentially off the Fed funds target according to the interest on reserves. While the risk of increases in the Fed funds rate may seem benign now, it may not be the case at some point in the future.

But the most important aspect is the duration risk that’s been removed from the market, since much of that would be marked to market had it remained in street portfolios, where as I say it becomes super sensitive to even small interest rate moves when rates in general are so close to the zero bound.

Cont’d...
January 5, 2012 8:42 AM

JKH said... ....

But my point really was about risk in broad terms. The Fed has removed risk from market portfolios, other things equal, and these market portfolios would have required equity capital allocation to bear such risk – and that includes equity capital allocated to support interest rate risk. So it makes a difference. And there have been issues in credit risk earlier on, etc.

And one of the important points here is that the Fed indeed has the capacity to remove gargantuan amounts of risk from the market, if it chooses, simply due to the (MMT) insight that it can fiat that transfer of risk with ease. And it can do this without having to worry about marked-to-market risk, because in all cases it has the inherent balance sheet strength to hold out until maturity of its assets.

And it doesn’t really use marked to market accounting for its bonds portfolios, for the most part, for that reason. And finally, when MMT makes the point that the Fed is removing interest income from the market, we should just bear in mind that there is a “risk/return trade off” involved. It’s also removing risk from the market. And the pricing of the transactions reflects that risk.

And the additional income the Fed has earned so far, while being income that has been removed from the market, is income that compensates the Fed for the risk it has removed from the market. And it was the sellers to the Fed that sought the removal of that risk from their portfolios.

The other thing I’d like to say is that my comment was quick and sloppy in a way (and this one is also quick), but to make a point. I do believe MMT is incredibly strong (as a competitor for academia) in its understanding of the monetary system and the banking system, and that’s critical to its essence – as per recent discussions.

My own personal view that MMT is “weak” in risk (and I might have added, weak in capital, in the sense that bank capital allocation is inextricably linked to risk management) is exactly that – a personal view on where MMT’s unusual strength in monetary operations and the subject of reserve management isn’t particularly well balanced with the other essential technical banking functions of risk management and capital management.

But I’d also like to say that I didn’t give MMT fair comparison against either MM or Austrian in that sense. If anything, MMT is probably still much better than those two in these areas of risk and capital management – and Warren Mosler is up to speed on this stuff of course, although I’m not sure how closely familiar he is with large bank operations in these areas.

Finally, you can measure the relative strength of MMT in monetary operations against the other two schools only if you go off the charts to do so, because there is just no comparison - and the subject matter is critically important to understanding economics more generally, in my view.

January 5, 2012 8:46 AM 

19 comments:

mike norman said...

What "risk" did the Fed remove when it bought Treasuries? It entirely removed income and not risk. And it removed the safest and most necessary collateral that exists, which also happens to be a collateral that the public continues to demand in the extreme.

Райчо Марков said...

Mike, I was about to ask this question.

Anonymous said...

Maybe JKH is talking about the purchase of other assets - not just Treasuries? QE also bought up some of the toxic junk, didn't it?

Brinn aka The Banker said...

I like this post and I believe that JKH is one of the most even-handed and knowledgeable posters in the MMT blogosphere. He appears to have intimate knowledge of both accounting and banking operations mixed with a thorough understanding of both heterodox and neoclassical economics. A real renaissance man.

What is your background JKH?

hbl said...

Since there is a new post and I only recently commented at the previous two posts, I'll move a copy of my [probably unread] comments here. In response to JKH's two comments listed above:

JKH,

Good comments! I replied to your comment at Perry Merhrling's (still awaiting moderation) before realizing there was a discussion and more detail here.

I largely agree with you here, but with two points I'd emphasize:

1. My personal view is that MMT is neutral on risk rather than weak on risk, and that individuals MMT authors treat risk in different individual ways (but maybe I've missed something)

2. I agree that the Fed's QE takes away interest rate risk from the private sector, but I think given the portfolio preference aspect of endogenous money that the effect is almost entirely borne by the financial sector, with the loss being a direct reduction in the net interest margin of the entire private financial sector (banks and non-banks). Whether this is "helpful" probably depends on the regulatory environment (how much forbearance is in play), the capital raising environment (something you allude to), and how this change in maturity transformation is shared between banks and non-bank financials.

My understanding is that after QE the non-financial private sector actually ends up with a slightly riskier asset composition (i.e., less treasuries relative to assets with credit risk) as it strives to regain its preferred portfolio duration mix.

hbl said...

My reply to JKH (pending moderation) on the original Perry Merhrling post was before I read his two more recent comments posted above, with which I mostly agree. But that older comment references the Fed's ability to perform a quasi-fiscal role in absorbing credit losses from the private sector, which I'm pretty sure I've seen MMT authors discuss:

-----
JKH,

I'm not sure I agree that neo-chartalism is inherently weak on risk, though perhaps certain of its writers have especially unconventional individual perspectives.

Also, I think whether "the true effect of QE is to take risk out of the market" depends on circumstances. If the Fed buys private sector assets such as [possibly toxic] MBS, then yes, it's like government throwing itself on a grenade to limit damage to the private sector, and probably even reducing the size of the underlying explosion in the process.

But if the Fed buys treasuries as it did exclusively during QE2, what risk is being removed, other than interest rate risk?

And as we've discussed previously in the context of a post I did, endogenous money includes not only that loans create "money", but also that the private sector chooses its own portfolio duration mix by using the financial sector's maturity transformation role as the tool to this end. Ramanan recently posted a crystal clear quote from Kaldor from 1982 on this dynamic.

So after QE the non-financial private sector reclaims the duration mix that it "wants", in effect undoing the effect of QE. This will in some respects ADD risk because the non-financial private sector's long duration assets will include a relatively lower percentage of risk-free government bonds than prior to QE, with additional private sector assets (that have credit risk) taking their place.

Райчо Марков said...

Dan Kervick mentioned toxic junk.

Isn't it illegal when the Fed buys toxic junk?
And how does the Fed distinguish good asset from junk?

Hugo Heden said...

@Mike -

Let me try to chip in with a preliminary answer --

It was not obviously not default risk the Fed removed by buying Treasuries.

But as I understand it, there is also a liquidity risk (I've locked up my money and won't be able to spend it if a profitable business opportunity turns up) and an interest rate risk (interest rates totally hike, and I'm stuck with my sucky low-rate bonds).

(This I got from JKH's "Thinking about risk in the broadest terms, you have such categories as credit risk, liquidity risk, interest rate risk, and foreign exchange risk.")

(Can't say I'm sure about this though.)

Tom Hickey said...

In QE 1 the Fed didn't just purchase tsys, and there were concerns that default risk could be involved. I am not sure of the time frame now. That may be before Fannie and Freddie were nationalized so the agency MBS became in effect explicitly govt liabilities instead only an implicit guarantee. But those questions about potential default were around then IIRC. The charge was that the Fed was absorbing toxic waste in exchange for reserves, which constituted a quasi-fiscal op.

Anders said...

JKH - you make another point which is that QE does not provide additional purchasing power, on the grounds that purchasing power is already available by repo.

Surely the monetarist rerort is as follows.

"It's irrelevant that I _could_ have secured additional purchasing power by repoing my Tsy. I hadn't repoed it, and my purchasing power pre-QE reflected the fact that I hadn't repoed it. Now, QE gave me more purchasing power and liquidity than I wanted, and as I don't want this extra liquidity, I decided to buy another risk asset to try and shed the excess liquidity."

Not sure of the response to this. Monetarists appear to have decent data attesting to the 'hot potato effect' for the asset management sector.

JKH said...

This is a very interesting subject for me and I wish I had the time to give it more justice right now. But just something quickly again on the Treasury bond risk issue:

Regarding the nature of risk on Treasuries, there was a classic book written in the early 1970’s called “Inside the Yield Book”. The entire book was applicable to the nature of risk in US Treasury bonds. It was written by two fellows from the New York investment bank Salomon Brothers, Sydney Homer and Martin Leibowitz. Warren Mosler would remember this book very well. The book was a must read back then for any serious trader of US Treasury bonds. I don’t think the book actually used the word “duration”, but duration was the concept that was inherent throughout the book. Duration measures the relationship between the cash flow structure of a bond to risk in the price movement of a bond, based on the level of interest rates and the potential for change in the level of interest rates. There are a number of interesting properties about duration:

a) The duration of a bond increases as term to maturity increases

b) The duration of a bond increases as the coupon rate (originally set) is lower

c) The duration of a bond increases as market interest rates decline

d) The price sensitivity of a bond increases as its duration increases

e) Duration and its relevance for the price sensitivity of a bond is one measure of the risk of that bond

Now put all that stuff together, and you have a fairly powerful statement about the nature of the risk that the Fed is taking out of the market when it swaps bonds for reserves. (The duration of reserves arguably is very short term, based on the interest rate sensitivity of the interest paid on reserves, which arguably will change whenever the Fed wants to change the funds target, which arguably it can do anytime it wants to.)

Now, you have to separate that entire discussion from another issue about risk, which is the way that risk works in a more general sense at the large institutional level.

Let’s take PIMCO as an example.

And let’s construct a make believe PIMCO scenario that is not entirely disconnected from historical fact, but that I’ve MASSAGED to a make believe status in total, just to make a point.

Suppose that PIMCO decided to sell a whole slug of Treasuries at what turned out to be the wrong time – i.e. just before interest rates declined.

You have to think about what’s going on at an institution like this when such a decision is made. That decision is the result of a strategy in terms of how they’re looking at the risk in the market in terms of interest rate scenarios, and the risk inherent in their own portfolio in terms of how its value would react in such various scenarios. In this case, they looked at the market and they looked at their portfolio, and they didn’t like what they saw.

Now, it is very important to visualize roughly what they saw in their portfolio.

Cont’d...

JKH said...

...

What they saw was their bonds, but they also their risk reports.

And what their risk reports would show in this case is, among other things, the price sensitivity of those bonds, based on a measure of duration (or based on some sort of mathematical transform measure of the duration concept, because the duration concept is at the heart of measuring all types of interest rate risk on bonds). And they would have come to a judgement (it turned out wrongly) that what they saw as the threat that rates might move back up was simply too much for them, given the heightened sensitivity of bond prices at such a low level of interest rates, and given that they saw better opportunities to deploy capital on a risk adjusted basis.

And so, in this adjusted make believe scenario, they sold their bonds, at what turned out to be the wrong time.

And the Fed bought them, at what turned out to be a good time for them.

But the point is not who won and who lost here; the point is how the world at large is viewing bond risk. And there are bulls and there are bears, but they all have risk reports, and they all decide on “risk on” or “risk off” at various points of time.

And given the reality of those risk reports out there, and given the reality that 21st century financial institutions allocate scarce capital to all types of risk – including interest rate risk – it is indisputable that the Fed has taken risk out of the market by purchasing Treasuries and offsetting them with reserves.

And finally, it’s very important not to get caught up in a fallacy of composition here. Because the fact that the liquidity of the Treasury market would have allowed even PIMCO to download its Treasury portfolio into that market doesn’t change the other fact that in doing what it did, the Fed pulled aggregate interest rate risk out of the market. Just sum up all the risk reports on Treasuries before and after the Fed’s extraction, and that’s the conclusion.

hbl said...

JKH,

Interesting details, thanks!

"it is indisputable that the Fed has taken risk out of the market by purchasing Treasuries and offsetting them with reserves"

Yes, that is the initial macro effect. However, if you agree with the endogneous money dynamic that RSJ and I have both written about and that I now see seems to be consistent with Kaldor (1982), then this initial effect is temporary. Remember, QE1 and QE2 had very little visible effect on levels or trends in broad money supply (such as MZM).

Bottom line (and granted this is *not* a dynamic that I've seen embraced by any MMTers, and as I said before I don't think MMT as a discipline either adds or detracts from best practices in macro risk analysis, it just seems to be [mostly] silent on them):

1. Post QE, the financial sector probably does have its interest rate risk reduced.

2. The non-financial sector gradually reclaims at a macro level the duration it "desires" via a menu of possible transformations of financial assets and liabilities, and is now forced to hold a higher percentage of private sector financial assets instead of risk-free treasuries. Thus QE most likely adds CREDIT risk to the private sector in places it did not exist before.

The simplest example for #2 (though I could give others) would be that banks replace some deposit liabilities with bond liabilities to satisfy the non-bank desire to restore its portfolio duration... now the non-banks hold bonds that include both credit risk and interest rate risk instead of treasuries that had only interest rate risk.

That said, I'm not sure whether it's possible to determine whether the net effect of the two would be a positive or negative for the private sector overall.

JKH said...

hbl,

The Fed's removal of risk allows the financial sector to redeploy risk capital in other ways . That's part of the intended effect of QE. It's stimulative as far as the financial sector is concerned, although it has nothing to do with lending because of additional reserves, of course.

Not sure about the non-financial sector. Doesn't seem that intuitive to me, but haven't really considered it.

hbl said...

JKH,

I'm not clear on how the financial sector would "redeploy risk capital in other ways". I understand that risk-weighted capital ratios might be improved as a short term effect, but I'm not convinced that capital is really any kind of non-temporary limitation to the financial sector's activities, especially when it comes to supporting genuine expansion of economic activity (funding businesses and such). I agree with Warren that capital is always available and it's only a question of price. Even if QE contributed to a lower price of capital, I'm not sure that necessarily drives the larger economy in any meaningful way. I would happily be corrected on this as I have no direct experience here.

Of course you might be correct that this was a policy GOAL of QE, I'm just questioning the actual effects.

By the way, I don't remember seeing others state it this way, but I think it's reasonable to say "loans [can] create capital" in the same way that "loans create deposits" -- there is just an extra step or two to get from deposits to capital. (i.e. - Borrower spends loan proceeds, a saver "downstream" in the circular flow of spending buys bank capital, essentially converting the transferred deposit to capital). But the point is capital should never be a lasting constraint on anything at the macro level, outside of a national financial panic / credit crisis.

In case you're curious when you have more time, here is the Kaldor quote I referred to (via Ramanan):

http://www.concertedaction.com/2011/12/24/merry-christmas/

and in here are a few graphics I put together reflecting one aspect of this endogenous money dynamic (scroll to the middle) though a more complete list of possible adjustment dynamics is at the end:

http://www.thoughtofferings.com/2011/06/visual-guide-to-endogenous-money-and.html

I should probably clarify these points with another post though, as the long duration assets held by the non-bank sector aren't shown explicitly in those graphics.

Игры рынка said...

"And the additional income the Fed has earned so far, while being income that has been removed from the market, is income that compensates the Fed for the risk it has removed from the market."

The Fed does NOT remove the interest rate risk from the market for which it has to be compensated. The Fed destroys this risk. There is no zero-sum game between the Fed and the market with regards to the risk-free yield curve and interest rates. The Fed itself IS the risk free yield curve and IS the interest rates. The Fed can not be a risk to itself by definition.

By the same token as the Fed destroys the risk, it frees up capital. However, it would be interesting to see a sizing exercise of this effect. But I do not know whether US banks are *regulatory* constrained by the interest rate risk in the banking book. It is a concept of Basel 2. Which is not to say that US banks do not measure this risk but it is a different story from being regulatory constrained by it.

JKH said...

Hbl,

“I agree with Warren that capital is always available and it's only a question of price.”

It’s not that simple. The fact is that banks are VERY sensitive to their cost of capital. Capital is treated as a scarce resource, unless a bank has measured its own position to be one of excess capital. It’s complicated. Most blogs are screaming that the US banking system is insolvent. If true, it would mean capital was very scarce and precious. Of course, the blogs are wrong. And that said, JP Morgan clearly believes it has excess capital, because it’s buying back shares – although doing so in part because it believes its shares are undervalued. Redeployment of capital may well occur in existing asset markets with risk, as much as new lending. There’s never been any guarantee that QE would affect new risk lending to a great degree. But there is a capital and risk reallocation effect.

Игры рынка

It’s a fact that QE removes interest rate risk from the market. It’s indisputable.

And compensation for the risk it removes is a fact. It’s called the yield curve. And it’s realized in the Fed profit remittances to Treasury.

And the Fed is exposed to net interest margin risk, depending on what it decides to do with its own liability pricing in the future.

hbl said...

JKH,

"Capital is treated as a scarce resource"

But how much is this just about maximizing return on equity for existing shareholders?

"Redeployment of capital may well occur in existing asset markets with risk, as much as new lending. There’s never been any guarantee that QE would affect new risk lending to a great degree. But there is a capital and risk reallocation effect."

Sorry, I'm unclear on your implications here. Supposing that QE reduces interest rate risk within the financial sector as discussed... Are you suggesting either/both/none of:

(1) This increases the bid for various assets other than treasuries, bestowing a positive wealth effect on the economy?

(2) This causes the financial sector to spend more on goods and services, or causes it to lend more to expanding businesses, either of which would boost GDP directly?

If neither of the above, then I'm not sure the significance of this aspect of QE beyond reducing *potential* losses to shareholders of the financial sector (while at the same time reducing their current interest income).

Игры рынка said...

"And compensation for the risk it removes is a fact. It’s called the yield curve. And it’s realized in the Fed profit remittances to Treasury."

This is what I do not agree with. The rest is semantics.

Remittances is about the joint machinery of Fed + Treasury. If you want to separate them and argue for political aspects then lets move one level higher in the "profit" hierarchy. Fed defines the yield curve by managing market expectations, Treasury creates assets with interest rate risk. These assets are by definition "loss" making to Treasury due to interest it has to pay. Everything the Fed "earned" or "lost" on these assets goes/belongs to Treasury. You can not say that the Treasury was compensated for the risk on its *own* assets/liabilities because there is no risk. It does not matter whether you have a middle man or not in between. There can be no risk and no compensation for it because these assets are not external to Treasury.

The only reason the interest rate risk exists in credit risk free assets is because Treasury issues interest yielding instruments and Fed sets the interest rates. So when issuing Treasury creates the risk. By redeeming Treasury destroys this risk. When purchasing Fed destroys this risk. When selling Fed creates this risk. And lets be clear what this risk is about. It is about income which flows from Treasury to the private sector. Obviously there can be no income which flows from Treasury to Treasury.

Finally, Fed does not *manage* this portfolio having interest rate risk or margins in mind. It even secured itself from political risk and got a "waiver" on its profits and capital should it decide to sell these assets at accounting loss.